The Map Equation framework.

A coding-theoretic approach to community detection: good communities are the ones that compress the description of a random walk best.

Since 2008, the framework has grown into Infomap (the reference algorithm), a family of visualizations, and ongoing research on higher-order, multilayer, and Bayesian community detection.

Start here

Explore

News

Latest releases & papers

All news

May 5, 2026

Release

Infomap v2.10

Per-level module counts in JSON output, idiomatic R package with SWIG bindings, library-safe no-output mode (changelog).

May 5, 2026

R Package

Infomap is now on R-universe

After many requests, Infomap is now available as an idiomatic R package with SWIG bindings. Install it from R-universe and run multilevel community detection directly from R — including a high-level cluster_infomap() helper that takes an edge list and returns modules, codelength, and a tidy node table. This brings the same algorithm and feature set as the Python API to the R community.

Feb 25, 2026

Release

Infomap v2.9

Enhanced CLI summary output, codelength correction for higher-order solutions, NetworkX multilayer graph state ID handling (changelog).

Feb 3, 2026

Publication

The best maps convey a great deal of information but require minimal bandwidth: the best maps are also good compressions.

M. Rosvall and C. T. Bergstrom, PNAS 105, 1118 (2008)